Sparse MoE models compressed using REAP (Router-weighted Expert Activation Pruning) method
-
cerebras/Qwen3-Coder-REAP-363B-A35B-FP8
Text Generation • 363B • Updated • 61 • 15 -
cerebras/Qwen3-Coder-REAP-246B-A35B-FP8
Text Generation • 246B • Updated • 323 • 21 -
cerebras/Qwen3-Coder-REAP-363B-A35B
Text Generation • 363B • Updated • 42 • 4 -
cerebras/Qwen3-Coder-REAP-246B-A35B
Text Generation • 246B • Updated • 32 • 7